AUGUST 15,1940
ANALYTICAL EDITION
453
0.5 p. p. m. of manganese (4) a t the iodide concentration TABLE 11. COMPARATIVE CHLORIXE VALUESWITH OI-SAPHTHOfound optimum for colorimetry (3) and employed here. There FLAVONE FILTRATION TECHNIQUE appears to be a concomitant' variation in the concentration KI G./ liter
pH
6.7
S.9
Chlorine Range Min. Max.
Apparent Chlorine i Actual Chlorine Concentration, R
20 0 . 2 4 20 0 . 7 5 1 0.22 1 0.50
0.32 0.96 0.80 0.81 0.24 0.93 0.70 0.96
0.85 1.07 1.00 0.96
1.08
0.96
1.19
20 0 . 2 8 20 0 . 8 0 1 0.28 1 0.70
0 . 2 8 0.89 0.90 0.83 0 28 0 . 7 9 0 . 7 0 0.81
(0.67)b 0.75 (0'70)h ( 0 . 7 2 ) b ( 0 . 7 2 ) b (0.5S)C
..
..
.. ..
0'92
..
:'
..
..
..
.. ..
.. .. .. .. ..
Mean R
1.01 0.94 0.95 0.96 0.89 0.79 0.79 0.81
20 0 . 2 6 0 . 2 6 0 . 8 5 .. (0.75)C 0.85 .. 0.68 20 0 . 8 0 0 . 9 0 0 . 6 3 0 : 7 3 .. (0:ii); (0'iS)c 0.55 1 0.27 0.28 0.54 0.50 1 0.72 0.80 0.69 0.78 .. (0.71) .. 0.74 a Actual chlorine value estimated by o-tolidine test under standard conditions. b Ammonia-acetic acid buffer used. None of parenthetic values enters mean:, C J% Ith second lot of barbital. None of parenthetic values enters mean. 10.0
TABLE111. CHLORISEEQUIVALEXT OF IODINE LIBERATED BY COLLoID.4L M A X G A X E S E DIOXIDE (With 20 grams of potassium iodide per liter of sample. 7 -
pH 6.7 8.9
0.0 P . p . 711. 0.02 0.00
Single observations)
Manganese Dioxide as P. p. m. Manganese 0.625 1.25 2.5 5.0 P. p . Wl. P . p . 711. P. p . lil. P . p . PI1. 0.02 0.00
0.03 0.00
0.07 0.00
0.15 0.00
10.0 P . p . ?i.i 0.50 0.00
of manganese used and consequent liberated iodine. The observations a t p H 8.9 agree with those made by the direct colorimetric test (3) in showing no interference.
Summary When potassium iodide and a-naphthoflavone are added to chlorinat'ed water, the iodine liberated is rapidly adsorbed from solution b y the colloidal indicator arid may be separated from wat'er-soluble colored matter by filtration. The precipitate may be dissolved in a volume much smaller than that of the original sample by passing alcoholic potassium iodide solution through the filter, which retains any insoluble material. The yellow iodine solution may be compared with permanent color standards calibrated against known concentrations of chlorine. This procedure is useful in the examination of highly colored seFages and trade Wastes, and in investigations of the effect of variations in pII and iodide concentration on the reactions of interfering substances. The addition of p-aminodimethylaniline hydrochloride for increased coloration is permissible, since interfering substances such as nitrite, oxidized iron, or manganese have been eliminated but t'his reagent is relatively unstable.
Literature Cited reagents in ube at the time and should include tests with known chlorine concentrations a t p H and iodide-ion concentrations covering the ranges to be employed. The method has been found useful in determining the chlorine demand of tannery wastes too highly colored for direct colorimetric or titrimetric tests. The oxidizing action of highly colored manganese dioxide sols (prepared in the manner previously used, 4 ) on potassium iodide was also studied, with results shown in Table 111; p H 6.7 represents the threshold for interference in about 2 minutes' reaction time with
(1) Am. Pub. Health dssoc., "Standard Methods for the Examination
of Water and Sewage", 8th ed., pp. 19-24, New York, American Public Health Association, 1936. (2) Byers, D. H., with Mellon, M. G., 1x1).ESG.C H m r . , Anal. Ed., 1 1 , 202-3 (1939). (3) Gilcreas, F. W., and Hallinan, F. J., J . Am. Water Works Assoc., 31, 1723-32 (1939). (4) Hallinan, F. J., and Thompson, W. R., J . A m . Chem. Soc., 61, 265-70 (1939). Time unit used fort was the minute (not stated in article). PRESEXTED before the Division of Water, Sewage, and Sanitation Chemistry a t t h e 98th Meeting of t h e American Chemical Society, Boston, Mass.
Analvsis of Black Enamels J
S. E. BERKENBLIT, Board of Transportation, City of New York, New York, N. Y.
T
HE analysis of black enamels and paints containing small proportions of pigment which consists entirely of
carbon presents considerable difficulty. Enamels of this type are used for painting passenger cars of the Independent Subway System operated by the City of Kew York-for example, Underframe Black enamel is applied on equipment under the body of the car, and Hand Rail Black enamel is used for painting the grab handle and step a t the end of the car and exposed galvanized iron parts. A Sign Box Black enamel is used inside the destination sign box, and a Striping Black enamel for painting a stripe between two different colors on a painted surface. These enamels usually contain from 2 to 4 per cent of pigment and 96 to 98 per cent of vehicle. The pigment consists entirely of carbon, either carbon black or lampblack, and the vehicle is a varnish containing synthetic or fossil gums. Considerable difficulty has been experienced in this laboratory in obtaining a clean-cut separation of the pigment from the vehicle and in determining the correct amount of pigment in the sample. Because of the small proportion of pigment present, a slight variation in the weight of the extracted pigment produces a considerable error in the final result.
Existing Methods All attempts to separate the pigment from the vehicle by the method of the American Society for Testing Materials (1)treating the sample with a standard mixture of solvents and centrifuging-proved unsuccessful, as the carbon remains in suspension and no amount of centrifuging will cause i t to settle to the bottom. Attempts to filter off the carbon were also unsuccessful, as it is so finely divided that it goes through the filter paper. I n addition it was found that the solvents employed in the above mixture precipitate the gums present in the vehicle. Subsequent treatments with additional solvents failed to redissolve the precipitated gums, as a result of which the pigment values obtained were as much as 100 per cent too high. Similar results were obtained by using the extraction mixture consisting of benzene, methyl alcohol, and acetone, recommended by Gardner ( 2 ) for blue and green paints. Unsatisfactory results were also obtained by following the method recommended by Gardner (W,p. 9.51) for determining the pigment in black baking enamels-1. e., diluting the sample liberally with benzene, settling, and filtering or centrifuging.
454
INDUSTRIAL AND ENGINEERING CHEMISTRY
I n addition to the above extraction mixtures, other solvents were used to obtain a complete separation of the pigment from the vehicle, such as methyl isobutyl ketone, Cellosolve (ethylene glycol monoethyl ether), and methyl Cellosolve, b u t in each case the separation of the pigment from the lrehicle was incomplete and the carbon invariably remained in suspension. The method developed by the writer gives a complete separation of the vehicle oils and resins from the pigment and eliminates the troublesome suspensions of the carbon in the extraction mixture. It has now been in use in the laboratory of the Board of Transportation of the City of New York for about 2.5 years and has given entirely satisfactory results.
Procedure Keigh from 6 to 7 grams of the paint into a weighed centrifuge tube of 60-cc. capacity. Add about 25 cc. of an extraction mixture, consisting of 4 parts of mineral s irits and 6 parts of petroleum ether, mix thoroughly x-ith a g i s s rod, add more extraction mixture to make a total of 50 cc., and centrifuge for 15 to 20 minutes at a speed of about 2000 r. p. m., or until the pigment has well settled. Decant the clear supernatant liquid and repeat the extraction twice with the same extraction mixture. This treatment will dissolve the oils and part of the gums and resins present in the vehicle of the paint. The balance of the gums will be precipitated and will thereby carry down the suspended carbon particles, leaving a clear supernatant liquid. After decanting t'he supernatant liquid, treat the gummy residue in the centrifuge t'ube with about 20 cc. of a 50-50 mixture of acetone and ethyl ether and stir thoroughly with the glass rod t o break up all the lumps. To this add sufficient of the mineral spirit's-petroleum ether mixture used above to bring the volume t o 50 cc. and centrifuge again for 15 to 20 minutes. Repeat the extraction two or three times with the acetone-ethyl ether solvent mixture followed by the mineral spirits-petroleum ether mixture, or until the extraction liquid appears free from color. These extractions will remove the greater part of the gums from the residue, but some will still remain occluded in the carbon. Transfer the entire residue of carbon and gums from the centrifuge tube into an Erlenmeyer flask and saponify wit'h 75 cc. of
VOL. 12, NO. 8
iV alcoholic potassium hydroxide under a reflux condenser for about 3 hours. Let stand for 1 hour to allow the carbon to settle as much as possible, then decant the alkaline liquid through a weighed and previously ignited Gooch crucible, and wash the residue in the crucible with a 50-50 mixture of alcohol and acetone to remove the last traces of gums and resins. Continue washing until all alkaline salts are washed out, dry the Gooch crucible in the air oven, and weigh. The difference between this weight and that of the crucible will give the weight of the carbon plus any mineral matter present. .4sh the carbonaceous residue in the crucible, cool, and weigh again. The loss in weight obtained will represent the weight of the carbon removed by ignition, which is calculated to percentage on the basis of the sample taken for analysis.
I n following this method i t is important to use the solvents in the extraction mixtures in the proportions indicated, as slight deviations from these proportions will make it impossible to break up the carbon suspensions. The ease with which the separation of the carbon from the vehicle in this type of enamels can be accomplished depends to a great extent on the nature of the gums and resins employed in the vehicle. With some enamels, such as those with a glyceryl phthalate base, it may be preferable to saponify the sample directly, omitting the preliminary treatment with solvents in the centrifuge. Results of analyses by the above method were found to check within 0.2 to 0.3 per cent.
Aclrnowledgment The author wishes t'o acknowledge with appreciation the valuable assistance of J. Boorstein and S . Utkovitz, mho carried out the experimented work in connection with this paper. Literature Cited (1) Am. SOC.Testing Materials, D21.5. (2) Gardner, H. A , , "Physical and Chemical Examination of Paints, Varnishes, Lacquers, and Colors", 8th ed., p. 966, W'ashington, Institute of Paint and Varnish Research. 1937.
Iron Content of Bread and Bread Ingredients CHARLES HOFFRIAK, T. R. SCHWEITZER, AND GASTON DALBY Ward Baking Company, New York, N. Y.
I
NCREASING attention is being paid to the mineral constituents of foodstuffs, and among these iron is one of the most important. Considerable work has been done on the determination of iron in various types of food, feeds, and biological materials. Discrepancies in the reported results may be due to the analytical method rather than to actual differences in total iron content, especially in the case of prepared foods such as bread. During a study of the iron content of a special cereal product i t was found that the sum of the iron contents of the ingredients was much greater than the iron content of the finished cereal. Since the colorimetric determination used was identical in all cases, the most reasonable explanation for the discrepancy was a loss of iron during ashing. I n vien of the fact that some iron salts volatilize even a t the lom-est ashing temperatures, a loss of iron in products of varied composition is not unexpected. Farrar ( 2 ) used calcium carbonate to prevent the volatilization of ferric chloride during the ashing process. Jackson ( 6 ) , in a critical study of ashing methods, found that the use of calcium carbonate mas not satisfactory, and recommended a wet-ashing method using nitric, sulfuric, and perchloric acids. I n an attempt to find a simple and accurate method for ashing, various procedures were studied in this laboratory, and
it was found that sodium hydroxide solution added to the sample before ashing prevented the loss of iron. A series of determinations in which the quantity of caustic solution was varied indicated the value of this modification of the ashing procedure. This method of ash treatment assures that all portions of the sample are in contact with the alkali. Samples covered with calcium carbonate before ashing may not be uniformly alkalinized. The use of nitric, sulfuric, and perchloric acids in a wet-ashing procedure will no doubt retain all the iron originally present in the sample, but the sodium hydroxide ashing method offers simplicity and ease of manipulation as well as apparent accuracy. After ashing, the thiocyanate colorimetric determination described by Winter (10) was followed. The method has been used in this laboratory for a period of years. The results have been consistent and have agreed excellently with bioassays for iron on wheat products n-here the availability of the iron is usually considered to be 100 per cent. Saywell and Cunningham (7) found close agreement between the thiocyanate and o-phenanthroline colorimetric procedures. Gray and Stone (4) compared thiocyanate and dipyridyl, and found that the two colorimetric methods gave close checks. Comparisons of the various colorimetric methods have not been made in connection with this study, as i t was felt that
